Most reasoning in artificial intelligence concerns higher-level functions, such as game playing, language processing, and symbolic problem solving. Common-sense reasoning is concerned with the understanding and manipulation of information about the everyday world of objects and their interactions. Deciding that a pile of objects is unstable or that a vehicle will not get through a red light in time are examples of common-sense reasoning. Unfortunately this is surprisingly difficult to automate as the problem domain is so ill defined and open-ended; for example, the closed-world assumption is invalid. One approach is to provide large amounts of domain knowledge (see CYC project), while others concentrate on reasoning about materials, physics, space, and time (see naive physics, imprecision). The topic has also stimulated research into philosophical and logical issues in a search for formal structures.